Impaired glucose tolerance and its co-variates among 2079 non-diabetic elderly subjects. Ten-year mortality and morbidity in the CASTEL study. CArdiovascular STudy in the ELderly.

This study evaluated the role of impaired glucose tolerance (IGT) as a risk factor in a general population of 2079 non-diabetic elderly subjects. The 10-year cardiovascular morbidity was similar in normal and IGT subjects. Mortality was greater in IGT, but the Cox equations of the hazard rate were different in younger and older subjects: age, sex, lung function (forced expiratory volume in 1 s, FEV1), serum uric acid, IGT and proteinuria were predictors of overall mortality in the age class 65-79 years, while only the first 4 were associated with cardiovascular mortality. The same four items also predicted overall survival in subjects over 79 years old, while only age and uric acid were predictors of cardiovascular mortality. In older subjects, total cholesterol showed an inverse predictive value. Hyperuricaemia (> 6.4 mg/dl) and proteinuria did predict mortality in normal but not in IGT subjects, while reduced FEV1 (< 60% theoretical) was predictive in all. In 65-79-year old subjects IGT predicted mortality provided that FEV1 was normal, while in those 380 years old IGT was not a predictor. These interrelationships should be taken into account to better understand the factors underlying mortality.